Você está na página 1de 5

PROCEEDINGS

THE INSTITUTION OF ELECTRICAL ENGINEERS

Control & Science

118

Direct design procedure for multivariable feedback


systems
F. Fallside, B.Sc, Ph.D., M.A., C.Eng., M.I.E.E., and H. Seraji, B.Sc.
Indexing terms:

Computer-aided design, Feedback, Multivariable control systems

Abstract
A direct design procedure is established for linear time-invariant controllable multivariable systems which
permits a straightforward calculation of the feedback matrix to attain prescribed closed-loop poles, it permits
the choice of the relative tightness of the feedback to each input, and hence allows design with feedback to
only some inputs (i.e. incomplete input feedback). Furthermore, it permits the design of the important
practical case where some states are not accessible. For these, the procedure produces the necessary and
sufficient condition for pole positioning to be possible, and permits the required feedback to be calculated.
The procedure is applicable to systems with any number of inputs and states. Computationally, the method
is extremely fast, with computation times of the order of seconds, making it well suited to computer-aided
control-system design.
Introduction
The theory of multivariable control systems is well
advanced, and several methods exist for choosing a feedback
law to achieve desired design objectives. In recent years, there
has been a considerable interest in choosing feedback laws for
controllable multivariable systems to achieve arbitrary
dynamics of the closed-loop system. Most of these techniques
make use of feedback from the state vector to assign the
closed-loop system poles to particular locations;1"4 others use
output feedback for the same purpose.5'6 The present paper
attempts to contribute to the design of state-feedback
controllers by extending a single-input design procedure to
the multi-input case.
The method proposed is based on the equivalence of the
closed-loop characteristic polynomials of a multi-input and a
corresponding single-input system. The latter is first designed
using a previously established direct design method for singleinput systems,7 and the result is then transferred back to the
former.
The method has a number of attractive features. It is
computationally very fast, with computation times of the
order of seconds, and is well suited to the computer-aided
design of control systems. It provides the designer with
complete freedom over the relative tightness of the feedback
to each input, and hence also allows design with feedback to
only some inputs, i.e. incomplete input feedback. A further
important feature is that it permits design with incomplete
state feedback, when some of the states are not accessible.
Since the method involves the use of the existing procedure
for single-input systems, this is first summarised.
1

Summary of single-input direct-design


procedure
Consider a controllable single-input system described

by

where x is the /?-state vector, u is the scalar input, and A and


b are constant matrices of appropriate dimensions. The
transfer-function representation of eqn. 1 is
X(s) = G(s)U(s)
(2)
where G(s) = (si - A)~lb

PROC. IEE, Vol. 118, No. 6, JUNE 1971

40 C&S12

(3)
x

where Gc(s) = (si- A + bp)~ b = {gi(s)}/H(s), i = 1, . . .,


n, is the x 1 closed-loop transfer-function matrix* from
V(s) to X(s) and H(s) is the characteristic polynomial of the
closed-loop system, H(s) = \sl A + bp\.
It has been shown that7
pi8i{s) = H(s) - F(s)

(4)

i.e. the scalar product of the feedback and the numerator


transfer function vectors is equal to the difference of the
characteristic polynomials of the closed-loop and open-loop
systems. This direct relationship between the feedback vector
and the closed-loop poles establishes a direct-design method
whereby the feedback required to shift the open-loop poles
to desired closed-loop positions can readily be calculated.
The feedback vector p is simply calculated by equating
coefficients of like powers of s in eqn. 4.
3

(1)

Paper 6432 C, first received 8th January and in revised form 17th
March 1971
Dr. Fallside and Mr, Seraji are with the University Engineering
Department, University of Cambridge, Trumpington Street, Cambridge
CB2 1PZ, England

i = ! , . . . , , is the

X(s) = Gc(s)V(s)

3.1

x = Ax + bu

= {gj(s)}jF(s),

n x 1 open-loop transfer-function matrix from U(s) to X(s),


and F(s) is the characteristic polynomial of the open-loop
system, F(s) = \sl A\.
If a feedback law u = v px, where v is the external input
and p is the -state-feedback vector, is now applied, the
transfer-function representation of the closed-loop system is

Direct design procedure for multi-input


systems
Procedure

Consider a controllable multi-input system described


by

x = Ax + Bu

(5)

* We note that the numerator polynomials of the closed-loop and open-loop


systems are the same8

797.

where x is the /i-state vector, u is the m-input vector, and A


and B are constant matrices of appropriate dimensions.
The design problem is to find the m x n state-feedback
matrix P such that the closed-loop system described by eqn. 5
and the feedback law u = v Px, where v is the external
m-input vector, has a prescribed behaviour characterised by n
given closed-loop system eigenvalues or closed-loop system
pole A,, . . . A/r
Now, the closed-loop system poles are the roots of the
characteristic equation
H(s) = \sl - A + BP\ = 0

(6)

If we set P = qp, where q is an m-column vector and p is a


vector, eqn. 6 can be rewritten as
\sl - A + Bqp\ = 0
or

\sl - A + bp\ = 0

(7)

where b = Bq is an (n X 1) matrix.
Comparison of eqns. 6 and 7 yields the following lemma.
Lemma: The closed-loop poles of a multi-input system which
has a plant matrix A, a control matrix B and a feedback
matrix P are coincident with those of an equivalent singleinput system which has the same plant matrix A, a control
matrix b and feedback vector/;, where b = Bq and P = qp.
Making use of this equivalence, the design problem can be
solved in three straightforward steps:
(i) Choose an m-vector q. In general, q is arbitrary except
for the special cases treated in Sections 3.3, 4 and 5.
(ii) Find the /z-feedback vector p required to position the
poles of the equivalent single-input system (A, Bq) at the
desired positions A,, . . ., An, using the single-input
direct-design procedure based on eqn. 4.
(iii) For the multi-input system (A, B) the required state
feedback matrix is P = qp.

3.3

Controllability

It is well known9 that arbitrary closed-loop pole


assignment can be achieved if, and only if, the open-loop
system is completely controllable. When the system of eqn. 5 is
completely controllable, the equivalent single-input system is
also completely controllable for an appropriate choice of q,
and the method can be used to calculate P.
In general, the choice of q is arbitrary so long as the pole
or poles to be moved to closed-loop locations remain controllable; i.e. they are not cancelled in the open-loop transferfunction matrix of the equivalent single-input system. This is
illustrated by an example in the Section 3.3.1.
When the system is only partially controllable, the equivalent single-input system is also only partially controllable for
any choice of q. In this case, an arbitrary dummy state feedback can be employed to make the system completely controllable,1 and the direct design procedure can then be used
to calculate the feedback matrix.
3.3.1

Example 2

Consider a controllable multivariable system described


by

2 3The design problem is to find the feedback 11= Px required


to place the poles at a, b, (a > 0, b > 0).
Choose

The method is illustrated by an example.


3.2

desired positions, and shows that P is dependent only on


qx\q2. This can be generalised, and a proof is given in
Appendix 9 which illustrates that, for a specified set of
closed-loop poles, the feedback matrix P depends only on the
relative magnitudes of the elements of q and hence is unique
for particular ratios between the elements of q.
The direct design method can readily be computerised,
and this has been done in a program which calculates P
from A, B, A,, . . ., A,, and q given as data input.

Example 1

Consider a controllable multivariable system described

q =

by

ro n

ri

Now, for the equivalent single-input system,

01

G(s)q = (si - A)~lBq


The design problem is to find the feedback u = Px required
to place the poles at 1 + j :
(i) Choose

P'l

s(s + 5)

3q2>-

F(s)
'
. . . .

'
(9)

Placing the poles at a, b implies that H(s) = (s + a)


(s + b) and thus, from eqn. 4,

Now, for theU2J


equivalent single-input system,

G(s)q = (si - A)~ Bq

= (s + a)(s
i =

thus,

(b - 5)
2q2) \_q2

and hence, assuming q2 ^ 0,

J
\2qx\q2

~,

2qx\q2 - 2q\\q?[\
^_x\q
,2 J
2~,~ 2q

2q2_

Note that, if qx + 2q2 = 0, the factor s + 5 is cancelled out


in eqn. 9, and hence the open-loop pole at s = 5 cannot be
moved (by feedback) to the closed-loop position s b.
Therefore qx + 2q2^0 guarantees control of the pole at
s= -5.

2qx\q2~\

(iii) Transferring back to the multi-input case,

= qp= \

s(s + 5)

(10)

foi + 2<h)
we obtain

px(qxs + q2) + p2(q2s) 0 2 + 2s + 2) - s2

2 -

+ b)-

5q2]

Two different cases are studied.


Case (/). To position the poles at 5 = 0 and s = b (i.e.
a 0, b 7*= 5), then, assuming

J, 2

(ii) Placing the poles at 1 j implies


H(s) = (s + 1 - j) (s + 1 + j) = s2 + 2s + 2 and
from eqn. 4,

-ti

(8)

On forming the closed-loop characteristic polynomial using


eqn. 8, H{s) = \sl - A + BP\ = s2 + 2s + 2, which is
seen to be independent of q. To complete the design, we
choose q; e.g. qx = 1, q2= 2.
Eqn. 8 gives an infinite choice of P to place the poles at the
798

Case (ii): To position the poles at s = 5 and s a,


(i.e. b 5, a ^ 0), then, assuming
92^0

(11)

we obtain
P

a
= qp = 7 -

"1

~q%\Vol.. 118,. No. 6,. J.UNE 1971


PROC. IEE,

Note that, if q2 = 0, the factor s is cancelled out in eqn. 9,


and hence the open-loop pole at s 0 cannot be moved (by
feedback) to the closed-loop position s = a. Therefore
#2 ^ 0 guarantees control of the pole at s 0. We conclude
that case (ii) cannot be achieved without feedback to u2.
4

Relative tightness of feedback to each


input and incomplete input feedback
It can be seen that, since

feedback), this is accounted for by setting the associated


element or elements of p to zero. For example, if there is no
feedback from the states xl and xm then we must set pt =
The question then arises as to the existence of an incomplete
feedback matrix P such that a set of closed-loop pole positions
A,, . . ., A,, can be achieved. Since the condition of open-loop
controllability for pole assignment9 (Section 3.3) only applies
to the case of complete state feedback, it is necessary to
answer the existence question by other means. This is
examined in Sections 5.1 to 5.3.1.

Q\P\--<l\Pn
P=qp =

[p\ / > , / ] =
Slml

5.1
UlmPl

QmPn.

. . . .

(12)

the relative magnitudes of the elements of q determine the


relative tightness of the feedback to each input. For example,
iffy, = Kc/j, the feedback to , is K times tighter than that to Uj.
It is thus appropriate to name q as the 'relative-tightness
vector'.
As a further consequence, if feedback to some input or
inputs is not desired or obtainable (incomplete input feedback), this can be treated by setting the associated element
or elements of q to zero. For example, if uk = w, = 0, set
Vk = 4i = 0.
In general, pole assignment with incomplete input feedback can be achieved so long as the conditions on q discussed in Section 3.3 are satisfied. (This was illustrated in
example 2. When a2 ^ 0 t n e conditions are satisfied, but not
otherwise.)
An example showing these two features is now given.
Example 3

4.1

1 1 -2

0 0"
01
.1 -2 0

2 0 -2 x + 0
12

1.

(13)

The design problem is to find the feedback matrix P which


positions the closed-loop poles at 1, 1, 2 with no feedback to u3 and with feedback to , five times tighter than that
to u2.
From the foregoing, a suitable choice of q is
(14)

q=
Then, for the equivalent single-input system,

5s2- Us + 14'
4s- 26

r -

The existence problem is dependent on the solution of


eqn. 4, and is best seen by an example.
5.1.1

Consider the controllable


1 -2"
-3 x +
x =
-6.
This has
G(s) =

Then, from eqn. 4,

+ 5s

or, in matrix form,


ax - 6
"0 1 0" 'P\
1 5 1 Pi = a2- 7
(18)
.4 0 2 .P3.
.3- 2.
If the system is completely controllable, the matrix on the
left-hand side of eqn. 18 is nonsingular, and hence

~P\
Pi
-P3.

0 1 0" -l
1 5 1
4 0 2.

6"
a2- 7
3 - 2.

"i -

or
10a, - 2a2 + a3 - 48 '

Pi(5s - ]l.s + 14) + p2(4s - 26) + p3(3s + 2s + 14)


= s3 + 4s2 + 5s + 2 - (s3 - 2s2 + 5.s + 8)
whence

1-23

097]

(15)

and
3-1

6-15 4-35
(16)
1-23 0-97
0
0
0
which has the required properties.
This calculation can be carried out using the computer
program mentioned in Section 3.2.

P = qp = 0-62

1
+ 6s2 + 7.y + 2

(.17)

On equating coefficients of like powers of s, there result the


three equations
p2 = a x - 6
Pi + 5p2 + P3 = a2 1
4/j, + 2/>3 = a3 - 2

3.s2 + 2 ^ + 1 4

H(s) = (s + l)(s + ])(s + 2) - .y3 + As2 + 5^ + 2

62

system
"0"
1
0

Considering, for the moment, complete state feedback


p = (p{p2p3) and, for arbitrary closed-loop poles, H(s) = s2 +
axs2 + a2s + 3, eqn. 4 reduces to

and

Example 4

Pi(s + 4) + p2(s2 + 5s) + p3(s + 2)


= s3 + axs2 + a2s + 3 - (.y3 + 6s2 + Is + 2)

Consider a controllable system described by


x =

Pole positioning with incomplete state feedback:


single-input case

Incomplete state feedback


It is seen from eqn. 12 that, if feedback from some
state or states is not desired or obtainable (incomplete state
PROC. IEE, Vol. 118, No. 6, JUNE 1971

2a, - 12
- 2 0 a , + 4a2 - a3 + 94

09)

If we now take incomplete state feedback with/;, = 0, eqn. 19


gives
10a, - 2a2 + a 3 = 48
(20)
This is the condition required on the closed-loop poles for
pole positioning to be possible.
The argument can be generalised as follows. Writing
eqn. 18 in a general form
Lpl = a a

UlJ

where L is an n x n matrix of coefficients derived from G(s),


p is the complete state feedback n-vector, a and d are the
n-vectors of the coefficients of H(s) and F(s), respectively
(excluding the coefficients of s") and T denotes transposition.
The condition for pole positioning with complete feedback
799

is thus det L # 0, which is always true if the system is completely controllable, and then, from eqn. 21,
pT = M(a - d)

(22)

system {A, Bq) has the transfer-function matrix


B

G(s)q = (si -

+ g

j3 + 9 j 2

33

(30, - 802)
(qx - 2q2)s
qxs2 + (60, - 2q2)s + (90, - 202;
L Qis2 + (20, + 2q2)s + (60, - 502) .

where M = L~l.
For incomplete state feedback, with r states not fed back,
r elements of pT must be set to zero and eqn. 22 then implies
(23)

Taking complete state feedback for the moment, eqn. 4


is reduced to

where a is the r x n matrix composed of the r rows of M


corresponding to the zero elements of pT. Taking p = ad,
eqn. 23 can be written as

P\{(<1\ - 2q2)s + (3qx - 8q2)} + P2{qxs2 + (60, - 2q2)s


+ (9qx - 2q2)} + p^{q2s2 + (20, + 2q2)s + (60, - 502)}

a(a - d) = 0

aa = p

= (s + 3) 3 - (s3 + 9s2 + 29s + 33)

(24)

. . . .
This imposes r conditions on the coefficients ax, . . ., an of
H(js) to be satisfied, and these, in conjunction with the
controllability requirement det L =j 0, constitute the necessary
and sufficient conditions for pole assignment with incomplete
state feedback. When these conditions hold, p can be calculated from eqn. 22.
Since the elements of a are related to the closed-loop poles,
eqn. 24 can also be written in terms of pole positions. In this
form, eqn. 24 implies certain forbidden regions for closedloop poles in the splane.
It can be noted that eqn. 24 can be interpreted as allowing
only /; r of the elements of a to be chosen arbitrarily, and
the remaining r elements of a are then given by eqn. 24. This
agrees with the complete freedom of choice of a for complete
state feedback when r = 0.
The condition of eqn. 24 can be readily computerised to
determine whether pole positioning is possible, and the feedback p can be calculated from eqn. 22. When a desired pole
pattern is unattainable, an attainable near set can be calculated by iterating on a to satisfy eqn. 24.
5.2

L(q)pT = a -

(25)

and eqn. 24 becomes


a(g)a =

(26)

The appearance of the elements of q in eqn. 26 in general


provides more degrees of freedom, and thus can ease the
existence problem of incomplete state feedback for multiinput systems compared with the problem for single-input
systems.
It is also seen that, if pole positioning is possible, i.e. if
eqn. 26 can be satisfied, q is no longer arbitrary, since ratios
of its elements are defined by eqn. 26. When these conditions
hold, P can be calculated from eqn. 25.
The multi-input problem can also be readily computerised
by starting on a or q to test for satisfaction of eqn. 26.
5.2.1

Equating coefficients yields


qx
0
qx - 2q2 6qx - 2q2
2>qx - 8q2 9qx - 2#2

Example 5

x =

1
2
8

1
-3
2

-2'
-2
-7

x+

"0 0"
1 0
.0 1

(27)

Find the state-feedback matrix P which positions the closedloop poles at 3, 3, 3 with no feedback from xx.
We take q = (qx q2)T, a r | d the equivalent single-input
800

P2\ =

6q2^0

0"
-2
-6

. (29)

which are the conditions for complete controllability, i.e.


det L = 0, we obtain
P\
Pi

1
q2 - 4qxq2 + 6qj

L/>3J

(30)

Setting/?, 0 gives the condition


qx + 202 = 0

. . .

(31)

which is the reduced form of eqn. 26 for this problem.


Accordingly, we form
1

P0

(32)

-01/0 J

\q2

using eqn. 30. Then, inserting the condition of eqn. 31, there
results
1 TO

'-9L0

-2

-41

(33>

1 2J

We note from eqn. 32 that the elements of q appear in ratio


form, and hence the solution for P is unique.
In the case where, for a particular a, eqn. 26 cannot be
satisfied for any q, pole positioning is impossible and the feedback matrix P does not exist. Eqn. 26, together with the
controllability requirement of exprs. 29, again constitute
the necessary and sufficient condition for pole positioning.
Pole positioning with incomplete state feedback
and incomplete input feedback
The direct design procedure also allows the analysis
of the problem when both the state feedback and the input
feedback are incomplete. The effect of this is that some of the
elements of both p and q are now zero. The analysis is
similar to that of Section 5.2, except that now there are less
degrees of freedom in eqn. 26 compared with the complete
case, since some of the elements of q are now' zero. The
design method is illustrated by an example.
5.3

5.3.1

Consider the system

42

which is eqn. 25 for this problem. Assuming that

Pole positioning with incomplete state feedback:


multi-input case

The existence of a feedback matrix P for pole positioning of a multivariable system (A, B) with incomplete state
feedback reduces to the existence of the feedback vector p
for the equivalent single-input system (A, Bq) with incomplete
state feedback. The analysis is thus similar in form to that of
Section 5.1.
However, now the elements of q appear in G(s)q, and hence
eqn. 21 becomes

(28)

Example 6
Consider a system described by
1 -1
0 -2
0
0
0
0

0
0
-3
0

0
0
4

x+

"0 0
1
1 0 -1
0
0 1
2
1
(34)
. . . .

The problem is to find the state-feedback matrix P which


PROC. IEE, Vol. 118, No. 6, JUNE 1971

results in a closed-loop characteristic polynomial H(s) =


s4 + axs3 + a2s2 + a^s -f- aA with no feedback to w3 and no
feedback from x4.
Accordingly, we choose q (qx q2 0)T. Then, for the
equivalent single-input system,
G(s)q = ( ^ >0~ ' # ? = ,

i m.

i -,w

i ^/

positioning, and, if these are satisfied, allows the calculation


of the required incomplete feedback matrix. The procedure
is applicable to systems with any number of inputs and states.
Since the procedure is straightforward, it is computationally
very fast, with computation times of the order of seconds and
hence is well suited to computer-aided control-system design.

i ^

(qxx + q2)(s + 2)(s + 3)


-qx(s -I- 3)(J
3)(s + 4)
4) + (q
3)(s + 4)
2)(* + 4)
J + 2)(j + 3 )
. . . .
(35)
On applying the single-input direct design procedure, and
assuming complete state feedback for the moment, eqn. 4
gives

8
1

px{-qx(s + 3)(s + 4) + to, + <72)(* + ) ^ + 3)}


+ /> 2 fai0+ D(J + 3)(J + 4)}

feedback systems', Proc. IEE, 1967, 114, (3), pp. 395-399


BRASCH, F. M., and PEARSON, j . B. : 'Pole placement using dynamic
compensators', IEEE Trans., 1970, AC-15, pp. 34-43

RETALLACK, D. c , and MACFARLANE, A. G. j . : 'Pole-shifting tech-

3)}
5

= ss44 + axs3 + a2s2 + a3s + o4


3

- (s + 10s + 35s + 505 + 24)


(36)

say. Equating coefficients of like powers of s in eqn. 36, there


results

= as2 + j3*2 + ys + 8

2tfi

_6^r2 -

6tf,

Uq2

P3

^ 2 )J

(37)

Uniqueness of feedback matrix P for particular


ratios of elements of q
Consider a choice of q = (qx . . . <7,,,)T which gives
p = (px . . . pn) for closed-loop poles at A,, . . ., A,,,. Then, for
the equivalent single-input system,

-PA-

Assuming that

+ 3q\ ^ 0

-Aq}

. . . .
(38)
which are the conditions for complete controllability, i.e.
det L ^ O , and setting/^ = 0 gives the condition
q_x
?2

niques for multivariable feedback systems', Proc. LEE, 1970,


117, (5), pp. 1037-1038
POWER, H. M., and PORTER, B. : 'Eigenvalue assignment in multivariable systems incorporating integral feedback', Electron. Lett.,
1970, 6, pp. 795-796
JAMESON, A.: 'Design of single-input system for specified roots
using output feedback', IEEE Trans., 1970, AC-15, pp. 345-348
DAVISON, E. j . : 'On pole assignment in linear systems with incomplete state feedback', ibid., 1970, AC-15, pp. 348-351
FALLSIDE, F., and SERAJI, H.: 'Design of optimal systems by a
frequency-domain technique', Proc. IEE, 1970, 117, (10), pp. 20172024
BROCKETT, R. w.: 'Poles, zeros and feedback: state space interpretation', IEEE Trans., 1965, AC-10, pp. 129-135
WONHAM, w. M.: 'On pole-assignment in multi-input controllable
linear systems', ibid., 1967, AC-12, pp. 660-665

Pi

^y2)

72 -

cc

~ P\

References
ANDERSON, B. D. o., and LUENBERGER, D. G.\ 'Design of multivariable

Acknowledgment

The authors wish to thank R. V. Patel of the University of Cambridge Computer Laboratory for collaboration in
developing the computer program referred to in Section 5.1.

qx(s
q2(s

1 6 4 a - 25ff - y + 55
167a - 19j3 - 3y + S ' * '

'

'

Appendix

and

assuming the denominator is nonzero, which is the reduced


form of eqn. 26 for this example. The condition for pole
positioning is thus 167a 19/? 3y + S =j 0.
For an attainable pole position, eqn. 39 defines qx\q2. On
solving eqn. 37 for/;, the feedback is given by

His) = is - A,) . . . (s - A,,)

Now/? = (px . . . pn) is defined by eqn. 4 as

His) - F(s) =

pigi(s)

(41)

and, for the multivariable system, P = qp. Suppose that all


the elements of q are now scaled by a constant p to give a new
(Pi

Pi

PT>

0)

"(-49a + 5 + y)r
1
10 -

2A-

-21)

49a + 5p + y
0

It is again seen that P depends only on the ratio qx\q2 of the


elements of q, and hence is unique. We see again that eqn. 26
and the controllability requirement of exprs. 38 constitute
the necessary and sufficient conditions for pole assignment.
The design technique for systems with both incomplete
input and state feedback can be computerised in a similar
manner to that of Section 5.2.

j8(5 - 2r) - y

P(S-2r)-y

- 2r) + y}

r{a(31 - 16r) -

0"

a(31

(40)

0J

q' = pq. Note that this does not affect the ratios of the elements of q. In this case,
1

and, from eqn. 4,

His) - Fis) =

Conclusions

A direct-design procedure has been presented for


calculating the feedback matrix to assign arbitrary pole
positions to a controllable multivariable system.
One of its features is that it allows the treatment of systems
with incomplete input and incomplete state feedback, which
often occur in practice. For the case of incomplete state feedback, it produces necessary and sufficient conditions for pole
PROC. IEE, Vol. 118, No. 6, JUNE 1971

W
(42)
/=!
Comparing eqns. 41 and 42, we concludep'i = - ph i = 1,...,
1
P
n, and hence p' = -p. Finally, for this choice of q,
P
P' = q'p' = (pq) (-pj = qp
(43)
which is unchanged. P is therefore unique for a particular
choice of ratios of the elements of q.
801

Você também pode gostar